6.828 lab1 Exercise 2 死循环问题

2020-04-21 11:59:06 +08:00
 dogedoge

https://pdos.csail.mit.edu/6.828/2018/labs/lab1/

版本:

The ROM BIOS 这个章节,A 窗口 make qemu-gdb,B 窗口 make gdb

A:

/usr/libexec/qemu-kvm -drive file=obj/kern/kernel.img,index=0,media=disk,format=raw -serial mon:stdio -gdb tcp::25000 -D qemu.log  -S
VNC server running on `::1:5900'

B si 单步调试:

[f000:fff0]    0xffff0: ljmp   $0xf000,$0xe05b
[f000:e05b]    0xfe05b: cmpl   $0x0,%cs:0x69c8
[f000:e062]    0xfe062: jne    0xfd0f7
[f000:e066]    0xfe066: xor    %dx,%dx
[f000:e068]    0xfe068: mov    %dx,%ss
[f000:e070]    0xfe070: mov    $0xf1399,%edx
[f000:e076]    0xfe076: jmp    0xfcf8c
[f000:cf8c]    0xfcf8c: cli
[f000:cf8d]    0xfcf8d: cld
[f000:cf8e]    0xfcf8e: mov    %eax,%ecx
[f000:cf91]    0xfcf91: mov    $0x8f,%eax
[f000:cf97]    0xfcf97: out    %al,$0x70
[f000:cf9b]    0xfcf9b: in     $0x92,%al
[f000:cf9d]    0xfcf9d: or     $0x2,%al
[f000:cf9f]    0xfcf9f: out    %al,$0x92
[f000:cfa4]    0xfcfa4: lidtw  %cs:0x69b8
[f000:cfaa]    0xfcfaa: lgdtw  %cs:0x6974
[f000:cfb0]    0xfcfb0: mov    %cr0,%ecx
[f000:cfb3]    0xfcfb3: and    $0x1fffffff,%ecx
[f000:cfba]    0xfcfba: or     $0x1,%ecx
[f000:cfbe]    0xfcfbe: mov    %ecx,%cr0

切入 PE 之后:

The target architecture is assumed to be i386
=> 0xfcfc9:     mov    $0x10,%ecx
=> 0xfcfce:     mov    %ecx,%ds
=> 0xfcfd0:     mov    %ecx,%es
=> 0xfcfd2:     mov    %ecx,%ss
=> 0xfcfd6:     mov    %ecx,%gs
=> 0xfcfd8:     jmp    *%edx
=> 0xf1399:     sub    $0x8,%esp
=> 0xf139c:     movl   $0xf4254,0x4(%esp)
=> 0xf13a4:     movl   $0xf390a,(%esp)
=> 0xf13ab:     call   0xee4dd
=> 0xee4dd:     lea    0x8(%esp),%ecx
=> 0xee4e1:     mov    0x4(%esp),%edx
=> 0xee4e5:     mov    $0xf4200,%eax
=> 0xee4ea:     call   0xedd5a
=> 0xedd5a:     push   %ebp
=> 0xedd5b:     push   %edi
=> 0xedd5c:     push   %esi
=> 0xedd5d:     push   %ebx
=> 0xedd5e:     sub    $0xc,%esp
=> 0xedd61:     mov    %eax,%ebx
=> 0xedd63:     mov    %edx,0x4(%esp)
=> 0xedd67:     mov    %ecx,%ebp
=> 0xedd69:     mov    0x4(%esp),%esi
=> 0xedd6d:     movsbl (%esi),%edx
=> 0xedd70:     test   %dl,%dl
=> 0xedd72:     je     0xedfb6
=> 0xedd78:     cmp    $0x25,%dl
=> 0xedd7b:     jne    0xede1b
=> 0xede1b:     mov    %ebx,%eax
=> 0xede1d:     call   0xec570
=> 0xec570:     mov    %eax,%ecx
=> 0xec572:     movsbl %dl,%edx
=> 0xec575:     call   *(%ecx)
=> 0xec565:     mov    %edx,%eax
=> 0xec567:     mov    0xf683c,%dx
=> 0xec56e:     out    %al,(%dx)
=> 0xec577:     ret

=> 0xede22:     jmp    0xedfaa
=> 0xedfaa:     lea    0x1(%esi),%eax
=> 0xedfad:     mov    %eax,0x4(%esp)
=> 0xedfb1:     jmp    0xedd69
=> 0xedd69:     mov    0x4(%esp),%esi
=> 0xedd6d:     movsbl (%esi),%edx
=> 0xedd70:     test   %dl,%dl
=> 0xedd72:     je     0xedfb6
=> 0xedd78:     cmp    $0x25,%dl
=> 0xedd7b:     jne    0xede1b
=> 0xede1b:     mov    %ebx,%eax
=> 0xede1d:     call   0xec570
=> 0xec570:     mov    %eax,%ecx
=> 0xec572:     movsbl %dl,%edx
=> 0xec575:     call   *(%ecx)
=> 0xec565:     mov    %edx,%eax
=> 0xec567:     mov    0xf683c,%dx
=> 0xec56e:     out    %al,(%dx)
=> 0xec577:     ret
...

中间隔开的那一段一直到 ... 就是死循环,但是如果这个时候 c continue 的话 A 又可以进 kernel,而且试了断点,并没有进 0x7c00

找到一篇:https://stackoverflow.com/questions/11408041/how-to-debug-the-linux-kernel-with-gdb-and-qemu/33203642#33203642,把 A 窗口换成直接执行:

/usr/libexec/qemu-kvm -drive file=obj/kern/kernel.img,index=0,media=disk,format=raw -serial mon:stdio -D qemu.log -S -s

其实就是换了 gdb TCP 1234 端口,然后按照 stackoverflow 上在 B 连 1234:

target remote localhost:1234

最终 si 还是进了死循环... 另外试了 -bios 参数,也确实是使用的 seabios.bin

已经查不动了,求大佬解答~

4709 次点击
所在节点   汇编
0 条回复

这是一个专为移动设备优化的页面(即为了让你能够在 Google 搜索结果里秒开这个页面),如果你希望参与 V2EX 社区的讨论,你可以继续到 V2EX 上打开本讨论主题的完整版本。

https://www.v2ex.com/t/664577

V2EX 是创意工作者们的社区,是一个分享自己正在做的有趣事物、交流想法,可以遇见新朋友甚至新机会的地方。

V2EX is a community of developers, designers and creative people.

© 2021 V2EX